<- Back to projects

AI Video Player

Desktop video player (NeuraPlayer) with AI subtitles, semantic search, summarization, tutor Q&A, translation, and media tools β€” built with Electron, React, Express, and FastAPI.

Electron React Express FastAPI Gemini FAISS FFmpeg
AI Video Player Screenshot 1
AI Video Player Screenshot 2
AI Video Player Screenshot 3
AI Video Player Screenshot 4
AI Video Player Screenshot 5
AI Video Player Screenshot 6
AI Video Player Screenshot 7
AI Video Player Screenshot 8
AI Video Player Screenshot 9
AI Video Player Screenshot 10
AI Video Player Screenshot 11

#Overview

AI Video Player is a full-featured desktop app for local libraries and YouTube/remote URLs. It layers AI on top of classic playback β€” auto subtitles in 100+ languages, semantic transcript search, summaries & chapters, a streaming tutor, translation, emotion analysis, quizzes, and FFmpeg-powered clip tools.

#Core Playback

  • πŸ“š Video library β€” folder scan, thumbnails, search & sort
  • πŸ”— URL playback β€” YouTube embed + direct media URLs via yt-dlp
  • ⏩ Player controls β€” A/B loop, 0.25×–4Γ— speed, PiP, fullscreen
  • πŸ“‹ Playlist queue β€” auto-advance with rebindable hotkeys
  • πŸ’Ύ Watch history β€” resume playback with HTTP range streaming

#AI Features

  • πŸ—£οΈ AI subtitles β€” AssemblyAI / Gladia / Whisper / Gemini routing with speaker diarization
  • ⏱️ Auto-sync β€” align cues to speech onset via FFmpeg
  • πŸ” Semantic search β€” FAISS + sentence-transformers, seek-on-click results
  • πŸ“ Summary & chapters β€” Gemini or local Ollama, pointwise or paragraph
  • πŸŽ“ AI tutor β€” streaming Q&A about video content (SSE)
  • 🌐 Translation β€” MyMemory β†’ Gemini/Ollama fallback
  • ✨ Smart highlights β€” key moments spaced across the timeline
  • 😊 Emotion / sentiment β€” per-segment tone analysis
  • ❓ Quiz generation β€” comprehension Q&A from the transcript

#Clips & Processing

  • βœ‚οΈ Clip creator, screenshots, named bookmarks
  • 🎞️ GIF creator (two-pass palettegen)
  • πŸ”— Clip merger, speed remapping, frame extraction
  • 🎯 Subtitle auto-sync for exported media

#Tech Stack

Desktop & Frontend

  • Electron β€” desktop shell
  • React + Vite β€” UI
  • Custom dark glassmorphic design

Backend

  • Node.js + Express β€” API
  • FFmpeg β€” media processing
  • yt-dlp β€” YouTube audio/video

AI Service

  • FastAPI + FAISS search
  • Gemini 2.5 Flash
  • AssemblyAI / Gladia / Whisper
  • Ollama β€” optional local LLM

#Architecture

ai-video-player/
β”œβ”€β”€ electron/       # Main process + IPC + service launcher
β”œβ”€β”€ frontend/web/   # React + Vite player UI
β”œβ”€β”€ backend/        # Express API (transcribe, tutor, clips…)
└── ai-service/     # FastAPI + FAISS + Whisper/Gemini